基于约束模板实现认知无线电网络频谱分配的遗传算法

摘要

对遗传算法用于解决认知无线电网络频谱分配中的无干扰约束条件问题进行分析,提出一种基于约束模板进行认知无线电网络频谱分配的遗传算法模型,该模型以系统效用最大化为目标求得无干扰的约束模板,利用约束模板的标号组合构成染色体,再通过遗传算子的进化选择适应度函数更大的约束模板组合,最终得到无干扰的可行分配策略.仿真结果表明,该算法可以避免利用传统遗传算法进行频谱分配的模型在进化过程中对种群中染色体重复性的无干扰约束处理,及由于无干扰约束处理引起的适应度函数减小等问题,可获得更高的系统效用.%This paper discusses the noninterference constraints of the genetic algorithm based spectrum assignment problem in cognitive radio networks, and proposes a model for spectrum assignment using the genetic algorithm based on constraints templates. The noninterference constraint templates are calculated aiming at maximizing the system utility , and the chromosome in the proposed algorithm is composed of the indexes of the constraint templates. Thus the combination of the constraint templates with high fitness is obtained through the evolution of genetic operators, and then the feasible spectrum assignment strategy without interference is achieved. The simulation results show that the proposed spectrum assignment model can improve the system utility and avoid some problems in conventional genetic spectrum assignment models, such as repetitive constraints-free process of the population and the corresponding reduction of fitness function caused by the constraints-free process.